Movie Gateway
Built a full-stack movie ticket booking platform with theatre selection, real-time seat availability, and booking confirmation using the MERN stack, reducing the user booking flow to under three steps.
KANISHK S.K
B.TECH AI&DS || SRI KRISHNA COLLEGE OF ENGINEERING AND TECHNOLOGY || FULL STACK DEVELOPER
Coimbatore, IndiaArtificial Intelligence and Software Development
2roles
28skills
2education
4credentials
About
AI and Data Science undergraduate with experience in machine learning, full-stack development, NLP, retrieval-augmented generation, and computer vision.
Experience
Machine Learning Intern
CodeAlpha
2025
Applied feature engineering and hyperparameter tuning on real-world classification datasets using Scikit-learn and Python, achieving a 15% improvement in model accuracy over baseline. Built end-to-end ML pipelines covering preprocessing, training, evaluation, and deployment-readiness.
Frontend Development Intern
AlfidoTech
2024
Designed and developed responsive, component-based frontend websites using ReactJS, improving UI consistency and reducing design iteration cycles through reusable components.
Education
SRI KRISHNA COLLEGE OF ENGINEERING AND TECHNOLOGY
BTech AIDS, Computer and Information Sciences and Support Services
Aug 2024 - Oct 2028
Sri Krishna College of Engineering and Technology
B.Tech, AI & Data Science
2024 – 2028
CGPA: 7.7
Skills
Full-Stack DevelopmentFront-End DevelopmentProblem SolvingWeb DesignWeb DevelopmentResponsive Web DesignPythonJavaScriptNLPTransformersRAGLangChainScikit-learnReactJSNode.jsDjangoMongoDBMySQLFirebaseGitPower BIGoogle CloudOpenCVFAISSLLMTF-IDFCosine similarityCompetitive programming
Projects
Career Connect App
Architected a full-stack student-to-company matching platform using ReactJS and Django REST API with ML-powered resume shortlisting via TF-IDF and cosine similarity, improving candidate match accuracy by 15% over keyword-based filtering. Reduced job-search time by eliminating manual tracking.
Crowd Detection System
Developed and trained a CNN-based crowd density estimation model using OpenCV on annotated surveillance datasets, achieving 98%+ head-count accuracy in dense environments. Implemented density-map regression for real-time inference in public safety monitoring systems.
RAG Document Analysis Pipeline
Engineered a retrieval-augmented generation pipeline for enterprise document QA using LangChain, FAISS vector store, and OpenAI API; achieved 95% response accuracy after embedding optimization and retrieval chunking improvements. Reduced hallucination by grounding responses in retrieved document context.